Mitogen activated protein kinases (MAPKs) are serine-threonine protein kinases that play critical roles in regulating cellular homeostasis. MAPKs are members of a three kinase relay system composed of the MARK, MARK kinase (MKK) and MARK kinase kinase (MKKK). Four MARK families, ERK1/2, p38, JNK and ERK5 are members of an integrated MARK signaling network. There are 7 MKKs but at least 20 MKKKs in the MARK signaling network. We have made gene knockouts to ablate expression or knockins to inactivate the kinase activity of four MKKKs (MEKK1, 2, 3, 4). Ablation or kinase inhibition of each MEKK results in a unique physiological phenotype in the mouse. Different MEKKs were found to regulate neurulation, vascular and placenta! development, and dissemination of tumor cells during metastasis, tissue remodelingfollowing ischemia or hypoxia, and specific aspects of chronic inflammation. The hypothesis of the current proposal is that MKKKs differentially integrate the activation of specific MAPKs in cellular responses to diverse stimuli. Despite the importance of MARK signaling networks in cell and animal physiology there is no comprehensive understanding of how different MKKKs coordinately regulate the MARK network for cellular homeostasis. We have shown that RNAi faithfully recapitulates targeted gene knockout of MKKKs in cells. Therefore,to define the function of MKKKs, an RNAi screen to ablate expression of a test set of 8 MKKKs that have particular relevance in control of homeostasis, will be performed in four human breast cancer cell lines. The aims of the proposal will define: 1. The functional consequence of MKKK knockdown on the signaling dynamics of the MARK network. 2. Role of MKKKs in the spatiotemporal activation of ERK1/2, JNK, p38 and ERK5. 3. To understand the transcriptional mechanism for changes in cytokine and protease expression resulting from specific MKKK knockdown. 4. To analyze the consequence of specific MKKK knockdown on orthotopic tumor cell growth and metastasis. 5. From these data a multivariate model based on Partial Least Squares will be developed to define the function of specific MKKKs in controlling signaling networks and physiological outputs. The creation of this systems-level model, along with iterative cycling between model and experiment, will allow us to define the function of specific MKKKs in controlling signaling networks related to tumor cell phenotype including metastasis-related physiological outputs.